The Apertis pipeline imports backport and/or proposed-updates versions
by default, this is not needed and leads to miss some security updates.
Because, importing versions from backports and proposed-updates is not
part of our standard workflow, this should only be optional to avoid
confusing the pipeline.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

`gbp import-dsc` puts all the new changelog entries in the commit
message, generating big walls of text when, for instance, importing a
new package for the first time.
GitLab then puts the whole log message in the `CI_COMMIT_MESSAGE` env
var, which makes the runner to fail during the executor preparation
section with the error:

    exec /usr/bin/dumb-init: argument list too long

See: https://gitlab.com/gitlab-org/gitlab/-/issues/392406



To avoid that, trim the message to only keep the first line.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

`flake8` said:

    tools/ade:1105:13: E741 ambiguous variable name 'l'

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

`flake` lamented:

    tools/ade:511:21: E722 do not use bare 'except'

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

`flake8` was complaining about:

    tools/ade:473:17: F841 local variable 'e' is assigned to but never used

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

`flake8` did report:

   tools/ade:288:26: W605 invalid escape sequence '\s'

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

`flake8` did report a typo:

    tools/ade:217:23: F821 undefined name 'user'

As far as I can tell nothing ever tries to stringify the `Device`
object, so that's probably dead code and that's why we never noticed.

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Signed-off-by: Emanuele Aina <emanuele.aina@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Since v2024dev2 the file pkg-config.multiarch is not longer shipped, so
in order to detect the sysroot architecture check the arch specific
libraries folder present.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Tool ade needs to find out release and distribution information for SDK and
sysroots. This was previously done reading /etc/image_version but this
custom file was dropped in favor of /etc/os-release.

In order to support both the old and the new approach fallback to
/etc/os-release if /etc/image_version is not found.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Walter Lozano authored 1 year ago and Emanuele Aina committed 1 year ago

Since same code is used in several places move it to new helper functions
to detect sysroots properties.

Signed-off-by: Walter Lozano <walter.lozano@collabora.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

If no mount for / (root) at given mount point can be established before mounting a sub partition, a virtual mountpoint
with a small tmpfs (ramdisk) is set up for / (root), so also non hierarchical images without a defined root partition
can be set up. This is usable, if partitions of an image are not mountable at one hierarchical mount tree.
Unmounting the image automatically removes the ram disk.

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Partprobe rereads disk structures to the kernel. After returning to caller the process is sometimes not completed.
So information about labels, part labels are sometimes are not available. A small sleep solves this issue so far, but
a better synchronization method would be more appopriate but too complex for such a small tool. And atm a reliable
algorithm is not clear.

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Thomas Mittelstaedt <thomas.mittelstaedt@de.bosch.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

When the soname changes, the name of the binary package changes
accordingly. In this case, ABI breakage is expected thus there is
no need to run abi-compliance-checker. apertis-abi-compare detects
this case and report the soname change without failing.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

This flag requires a string explaining the reason of importing
a package. This reason will be written in debian/changelog and
will be useful for a future reconsideration of the value of this
package in Apertis.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

apertis-abi-compare relies on sudo and fails if sudo
is not available.

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>

Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>